Lessons from the infuse trials: do we need a classification of bias in scientific publications and editorials?

Abstract

The original 13 Food and Drug Administration industry-sponsored recombinant human bone morphogenetic protein-2 (rhBMP-2) trials investigating its use in spinal fusion all reported no associated adverse events. However, subsequent series of studies began reporting complication rates that were much higher than those that were initially published. Critical analysis of the original rhBMP-2 industry-associated data found systematic alignment favoring positive outcomes with no proven clinical advantage over bone graft. The sources of potential bias leading to inaccurate reporting of original rhBMP-2 efficacy and safety profile include flawed study design, methodological technique, data reporting and analysis, and significant financial conflict of interest. As such, to ensure the integrity of the scientific literature, further measures should be taken by researchers, surgeons, authors, journal editors and reviewers to assess for potential sources of bias.